Uploaded image for project: 'JDK'
  1. JDK
  2. JDK-4144775

java.lang.OutOfMemoryError: out of memory allocating a JNI global ref

    XMLWordPrintable

Details

    • 1.2beta4
    • sparc
    • solaris_2.6
    • Not verified

    Description


      arthur.frechette@East 1998-06-02

      Baselevel Tested: JDK1.2Beta4-H
      Platform Tested: Solaris

      The following regression is occuring when running RMI scalability tests. This
      bug was not seen in the previous build (JDK1.2Beta4-G). The following stack
      trace is occuring on the RMI Test Server:

      java.lang.OutOfMemoryError: out of memory allocating a JNI global ref
      at java.security.AccessController.beginPrivileged(Native Method)
      at sun.rmi.transport.Transport.serviceCall(Transport.java:127)
      at sun.rmi.transport.tcp.TCPTransport.handleMessages(TCPTransport.java:584)
      at sun.rmi.transport.tcp.TCPTransport.run(TCPTransport.java:488)
      at java.lang.Thread.run(Thread.java:472)
      java.lang.OutOfMemoryError: out of memory allocating a JNI global ref
      at java.security.AccessController.beginPrivileged(Native Method)
      at sun.rmi.transport.Transport.serviceCall(Transport.java:127)
      at sun.rmi.transport.tcp.TCPTransport.handleMessages(TCPTransport.java:584)
      at sun.rmi.transport.tcp.TCPTransport.run(TCPTransport.java:488)
      at java.lang.Thread.run(Thread.java:472)
      java.lang.OutOfMemoryError: out of memory allocating a JNI global ref
      at java.security.AccessController.beginPrivileged(Native Method)
      at sun.rmi.transport.Transport.serviceCall(Transport.java:127)
      at sun.rmi.transport.tcp.TCPTransport.handleMessages(TCPTransport.java:584)
      at sun.rmi.transport.tcp.TCPTransport.run(TCPTransport.java:488)
      at java.lang.Thread.run(Thread.java:472)
      java.lang.OutOfMemoryError: out of memory allocating a JNI global ref
      at java.security.AccessController.beginPrivileged(Native Method)
      at sun.rmi.transport.Transport.serviceCall(Transport.java:127)
      at sun.rmi.transport.tcp.TCPTransport.handleMessages(TCPTransport.java:584)
      at sun.rmi.transport.tcp.TCPTransport.run(TCPTransport.java:488)
      at java.lang.Thread.run(Thread.java:472)


      The test scenario mixture was as follows:

      RMI Test Server: Total client connections expecting: 30
      RMI Test Client: 5 thread connections
      RMI Test Client: 5 thread connections
      RMI Test Client: 5 thread connections
      RMI Test Client: 5 thread connections
      RMI Test Client: 10 thread connections

      Total Client Connections to RMI Test Server: 30

      This regression bug occurs regardless of the following:

      o running with JIT
      o running without JIT
      o running with heap size set to -Xmx64m

      The above specific test scenario passes on the JDK1.2Beta4-G build with no
      problems.

      The above test succeeded on Beta4-H for client runs of 1,5,10,15,20,25. The
      cutoff point when the failure started occuring was with 30 clients.



      Attachments

        Activity

          People

            rschemersunw Roland Schemers (Inactive)
            afrechet Art Frechette (Inactive)
            Votes:
            0 Vote for this issue
            Watchers:
            0 Start watching this issue

            Dates

              Created:
              Updated:
              Resolved:
              Imported:
              Indexed: